A Proposed Framework for Face Recognition using Enhanced Local Binary Pattern Algorithm with Chinese Remainder Theorem

Abstract

One of the biometric methods that have recently gained attention across the globe is Face Recognition. This was due to the availability of practicable technologies, including movable results. Several studies have been carried out on face recognition for decades, but the problem is still largely unsolved. Significant progress has been made recently in this area as a result of advancements in face modeling and analysis techniques. While a system has been developed for face recognition, the problem of high processing time is still largely unresolved. This research framework proposed an Enhanced Local Binary Pattern (ELBP) algorithm for face recognition. The Local Binary Pattern (LBP) algorithm is a method used in facial feature dimensionality reduction. Standard LBP had challenges of computational complexity. Therefore, the LBP will be enhanced with the Chinese Remainder Theorem (CRT) and will be used for feature extraction to reduce computational time, Chicken Swarm Optimization (CSO) will be used for feature selection and classification will be done using a Support Vector Machine (SVM). Performance Evaluation of the system will be done by comparing the computation time result obtained from the combination of LBP-CSO and ELBP-CSO. The ELBP-CSO is expected to have a lower computation recognition time than the LBP-CSO.